Learning Item-Attribute Relationship in Q-Matrix Based Diagnostic Classification Models

TL;DR

This paper introduces a new estimation procedure for the Q-matrix in diagnostic classification models, enabling more accurate modeling of item-attribute relationships in cognitive assessments.

Contribution

It proposes a statistically principled method for estimating the Q-matrix and model parameters, with theoretical guarantees and applications to hypothesis testing and model selection.

Findings

The method has desirable large sample properties.

It facilitates hypothesis testing in diagnostic models.

The approach improves the accuracy of item-attribute relationship estimation.

Abstract

Recent surge of interests in cognitive assessment has led to the developments of novel statistical models for diagnostic classification. Central to many such models is the well-known Q-matrix, which specifies the item-attribute relationship. This paper proposes a principled estimation procedure for the Q-matrix and related model parameters. Desirable theoretic properties are established through large sample analysis. The proposed method also provides a platform under which important statistical issues, such as hypothesis testing and model selection, can be addressed.